Which way is better to reset index in a pandas dataframe?

Question

There are 3 ways to reset DataFrame index:

• reset_index(),
• reset_index(inplace=True),
• Manually setting the index via: df.index = list(range(len(df)))

Since inplace is going to be deprecated in pandas v2.0, which way is better, and why?

Answer 1

When assigning to the index, the rest of the data in your DataFrame is not changed, just the index.

If you call reset_index, it creates a copy of your original DataFrame, modifies its index, and returns that. You may prefer this if you're chaining method calls (df.reset_index().method2().method3() as opposed to df.index = ...; df.method2().method3()), but for larger DataFrames, this becomes inefficient, memory wise.

Direct assignment is preferred in terms of performance, but what you should prefer depends on the situation.